postcodes

Table of Contents

1. Overview
2. Usage
   • extract
   • prepare
   • pyramid
3. Datasets
4. Development
5. License

Overview

Demo

This project provides tools and algorithms to extract, prepare, and pre-compute German post code areas for prefix-based iterative lookups. As an example, say the user enters the German post code "10247" for Berlin, Friedrichshain.

When the user enters

• "1" we already want to pan and zoom the map to the Berlin / Brandenburg zone
• "10" we already want to pan and zoom the map to the Berlin region
• "102" we already want to pan and zoom the map to the Friedrichshain area

These iterative prefix-based lookups require pre-computation to efficiently pan and zoom the map in an instant.

Note: our initial use case is German post codes, but it's possible to extend it to other prefix-based post code schemes.

• https://en.wikipedia.org/wiki/Postal_codes_in_Germany
• https://wiki.openstreetmap.org/wiki/Free_The_Postcode#DE

Usage

The following shows command line tools to process and work with the post code data.

extract

To extract post codes from OpenStreetMap

postcodes extract postcodes.geojson berlin-latest.osm.pbf

This will generate a postcodes.geojson file with

• one MultiPolygon feature per post code area
• a postcode property on each feature with the area's post code

prepare

To prepare post code areas for lookup

postcodes prepare bounds.geojson postcodes.geojson

This will generate a bounds.geojson file with

• one Polygon feature representing the area's bounds per post code feature
• a postcode property on each feature with the area's post code

pyramid

To build a post code prefix search tree pyramid

postcodes pyramid pyramid.geojson bounds.geojson

This will generate a pyramid.geojson file with

• one Polygon feature representing the area's bounds per post code prefix feature
• a postcode property on each feature with the area's post code prefix

We generate five prefix tree levels (German post codes are made up of five characters)

• the first level contains bounds for all post code areas starting with a single character
• the second level contains bounds for all post code areas starting with two character

and so on.

Datasets

We recommend using OpenStreetMap snapshots provided by the lovely Geofabrik folks

wget http://download.geofabrik.de/europe/germany/berlin-latest.osm.pbf

For more see http://download.geofabrik.de

Note: OpenStreetMap data is licensed under the Open Database License (ODbL).

Development

Create a self-contained docker environment and enter it

make
make run

or use the image directly

docker run robofarm/postcodes --help

License

Copyright © 2020 robofarm

Distributed under the MIT License (MIT).